# Reputation Management RFP and Vendor Scorecard A reputation management RFP and weighted vendor scorecard: gates, scoring anchors, questions on assured results, and a worked example with its arithmetic. --- A reputation management RFP is a written request that makes every vendor answer the same questions on method, evidence, confidentiality, reporting and terms. A vendor scorecard turns those answers into weighted scores, so the choice rests on what each firm can show, not on who pitched best. The scorecard has three parts: gates that disqualify a vendor outright, weighted criteria scored from 0 to 5, and a verification step that checks every claim against a primary source. This guide is for a company, a family office or an adviser who must choose between several reputation firms and defend the choice. It covers the part before the brief and the contract: building the request, scoring the answers and catching the answers that should not be scored. The worked example is invented, with illustrative vendors, weights and scores, and every figure in it is computed from the stated assumptions. The FTC and Google rules cited were read at their source pages for this guide. They change, so check them before you rely on them.

The short version

Set gates before weights: a vendor that promises an outcome it does not control, or will not describe its method in writing, is out whatever its total.
Score seven criteria from 0 to 5, weight them to total 100, and put the largest weight on method and policy compliance.
Ask questions that force a concrete answer, then verify the answer: a policy page, a rule text, a reference call.
Treat a gap of less than 5 points out of 100 as a tie, and break it with reference calls and a small paid first phase.
Score contract terms last, as evidence of how the vendor behaves when the deal goes wrong.

## Why a scorecard beats a pitch day A scorecard settles one thing: which vendor wins on the criteria you chose before you heard any pitch. Pitch days reward the vendor with the best slides, and in reputation work the best slides often belong to the vendor making the biggest promise. Three failures repeat in unstructured selections. Anchoring: the first vendor you meet sets the price and the vocabulary. The assured outcome: a vendor names a date for a result it does not control, and the promise becomes the decision. The unweighted discussion: a charming account manager offsets a vague answer on method because nobody wrote down which matters more. A written rubric fixes all three. Weights are agreed before the first call, every vendor gets the same questions, and each score has a written reason a second reader can challenge. For a family office or a board, that file is also protection if a vendor later turns out to have used tactics that breach a platform rule. ## The seven weighted criteria and what each score means Seven criteria cover what decides an engagement. The weights below suit a matter involving a person or a family; adjust them before any vendor responds, never after. Score each criterion from 0 to 5. Weighted points are weight times score divided by 5, so a 5 earns the full weight, a 3 earns 60 percent of it and a 0 earns nothing. Totals run from 0 to 100.

Criterion	Weight	A 5 looks like	A 0 or 1 looks like
Method and policy compliance	25	Written method per tactic, each tied to a platform or search policy, with the limits stated	"Proprietary process", assured outcomes, or paid placement described as organic
Verifiable evidence	15	Reference calls with similar matters, and results shown with dates and the starting position	Logos, testimonials with no contact, or before and after images with no dates
Fit to the matter	15	Experience with your exact mix: removal routes, search, AI answers, your jurisdiction, your type of subject	One product for every client; no questions asked about your case
Confidentiality and data handling	15	Mutual NDA before the case file, named staff with access, deletion at the end	Case details requested before any agreement; shared inboxes; tool vendors undisclosed
Reporting and measurement	10	Baseline first, named metrics, a fixed report rhythm, raw evidence included	A single score from a private dashboard; no baseline
Commercial terms and exit	10	Scope-based fees, a clear exit, work product and accounts handed over	Long lock-in, fees tied to rankings they do not control, no handover
Team and accountability	10	Named lead and named people who do the work; escalation path	Sales team only; work done by unnamed subcontractors

Method carries 25 because it predicts the failure you cannot see. Weak reporting is annoying; a method that breaches a search or review policy can damage the subject's name further, later. Confidentiality carries 15 for a family or an executive because the vendor will hold the facts the subject wants private. A listed company or consumer brand with little to protect can move 5 points from confidentiality to reporting. Write a one-line anchor for 5, 3 and 1 on each criterion before scoring. Two readers should land within one point of each other. When they differ by two or more, the answer was unclear, and the fix is a follow-up question to the vendor, not an average. ## Gates come before weights: three reasons to remove a vendor A gate is a pass or fail test applied before totals are read. A vendor that fails a gate is excluded even if its weighted total is the highest, because a high total can hide a single disqualifying fact. The first gate catches tactics that breach a rule. The second catches claims the vendor cannot control: Google's page on removing information from Search says it can only remove things from Google Search results, and that the page stays on the host site and can still be reached through links, social media or other search engines. The third catches secrecy, which usually means one of the first two. A gate is not always a verdict on the firm. A salesperson may overstate what the delivery team would never do. Give the vendor one written chance to restate its position; if the restated answer is vague, the gate stands. ## What goes into a reputation management RFP, and the timeline The RFP is a short document, usually four to eight pages, that every vendor receives on the same day. Its job is to make answers comparable, so it asks structured questions and fixes the format of the response.

Building and running the RFP in seven steps

Define the matter in neutral termsOne page: subject, jurisdictions, types of content and what a good outcome looks like. Leave out identifying names until an NDA is signed.
Set the gatesPut the three gates in the RFP itself, so every vendor knows an assured outcome ends the process.
Fix criteria, weights and anchorsAgree weights and anchors before sending and record who scores. Never change weights after responses arrive.
Send the same questions to every vendorUse the question set below, a fixed answer template and one deadline. Allow one shared round of clarifying questions.
Score independently, then togetherEach scorer fills the sheet alone, with a written reason per score. Resolve gaps of two points or more with a follow-up to the vendor.
Verify before you shortlistCheck the claims behind the highest scores against primary sources and reference calls. Scores can go down here, and should.
Run a paid first phaseFor a close result, buy a limited baseline audit from the top two, same scope, and score the outputs.

A realistic process usually runs about six weeks. A fast-moving matter can compress it, but verification should not be the step that is cut.

Week	Activity	Output
1	Matter summary, gates, weights, anchors, NDA template	Locked rubric and RFP
2	RFP sent; clarifying questions answered for all	Written responses due at end of week
3	Independent scoring; gate decisions; follow-up questions	First scorecard
4	Verification and reference calls	Adjusted scores
5	Short presentations from the top two or three	Final scores
6	Paid first phase if tied; contract terms scored and negotiated	Decision memo

## Questions that expose assured results and black-hat tactics The questions that matter most are the ones that make a vendor commit to a concrete, checkable answer. A vague answer scores low on its own, so you do not need to catch a lie, only to notice the absence of detail. Ask each question in writing and keep the answers. The table pairs each question with an answer that passes and one that fails.

Question	An answer that passes	An answer that fails
What will you commit to, and what depends on others?	Commits to work, reporting and timelines for actions; names who decides each outcome	Commits to a result owned by a publisher, a platform or a search engine
List every tactic you would use, and the policy each one respects	Tactics named, each with its limit and risk	"We use proven methods"; no policy named
Will any link, article or review be paid for, and how is it labeled?	Paid items disclosed and marked as the platform requires	Placements described as organic that are bought
How do you obtain reviews for a client?	Invitations to real customers, no incentive tied to sentiment	Panels, "seeding", staff or relatives, or rewards for a rating
What do you do about reviews that are negative but real?	Respond, resolve, or dispute only where a platform rule is broken	Threats, offers of payment to remove, or "we remove anything"
Which sites do you own or control that publish about clients?	Any owned site is disclosed and labeled as ours	Review or ranking sites run by the vendor that present as independent

### The follow-up that separates good from rehearsed After any answer about method, ask: "Show me the policy page that allows this." A competent vendor opens the page. A vendor who deflects to experience has told you what you need, and method scores 2 or less. A firm willing to say what it will refuse to do for you is usually a firm that will refuse a shortcut that could hurt you. ## How to verify claims against primary sources A claim is verified when you have read the rule or page it depends on, not when the vendor has cited it. The table lists the checks that matter most in this field. The sources are the ones read for this guide, and each is a regulator or platform page.

Vendor claim	Primary source to read	What it settles
"We can build you five-star reviews"	FTC Consumer Reviews and Testimonials Rule, 16 CFR Part 465	Whether the method is a prohibited practice in the United States
"We place articles that link to your site"	Google spam policies, link spam section	Whether paid links must be marked, and what counts as link spam
"We will get this removed from Google"	Google's page on removing information from Search	What Google removes, and what only the host site can remove
"We can take down negative reviews for a fee"	Google Maps user contributed content policy	Whether payment to change or remove reviews is allowed

### What the FTC rule says about reviews The rule is 16 CFR Part 465, a US federal regulation on consumer reviews and testimonials. As read at the eCFR, it covers fake or false reviews, buying reviews conditioned on a sentiment, undisclosed insider reviews, company-controlled sites posing as independent reviewers, review suppression through unfounded threats, and buying or selling fake social media indicators where the person knew or should have known they were fake. The FTC's staff question and answer page says courts can impose civil penalties for knowing violations but states no amount, and that agencies and review management firms can be liable themselves. It matters for buyers because a business can also be liable for its own conduct, such as buying reviews it knew or should have known were false. The page gives warning signs for a hired source: reviews that arrive too quickly after purchase, in unusually large numbers, or that refer to the wrong product. A vendor offering a volume that fits those signs is the warning. The staff guidance is not definitive, your counsel decides how it applies, and the rule is US law: rules elsewhere differ and were not read for this guide. ### What Google's spam policies say about links Google's spam policies define link spam as creating links to or from a site primarily to manipulate rankings. The page lists buying and selling links, excessive link exchanges, paid articles, guest posts or press releases that carry links passing ranking credit or keyword-heavy anchor text, and automated link creation. Paid links are acceptable when marked nofollow or sponsored, and violating sites may rank lower or not appear. It also defines scaled content abuse as many pages created primarily to manipulate rankings. Two tests follow. A vendor selling "authority links" should say whether they are marked. A vendor that mass-produces pages to push results down should explain why they are not scaled content abuse. ### What Google says about review incentives Google's Maps contributed content policy says merchants may not offer incentives such as payment, discounts or free goods and services for a review, or in exchange for a revised review or the removal of a negative one. Google's Business Profile restrictions help page names what can follow: blocked new reviews for a set period, hidden existing reviews and a warning on the profile. ## Contract red flags, scored as evidence of behavior Score the draft contract last and feed it into "Commercial terms and exit", and back into method where it contradicts the RFP. Your counsel reviews the contract itself; here it is evidence of behavior. - A clause letting the vendor change tactics without notice, which defeats the method you scored. - A fee tied to a ranking, removal count or rating the vendor does not control. It rewards the shortcuts the gates exist to stop. - Content, domains or profiles staying with the vendor after the term. - A long minimum term with no exit if a gate fact comes to light. - Confidentiality that binds you but not the vendor's subcontractors. If a clause contradicts the RFP, hold the vendor to the RFP answer and lower the method score by a point. ## A worked scoring example, with the arithmetic The scenario is invented. A family office needs help with accurate but unflattering search results about a principal and receives three responses. Weighted points are weight times score divided by 5, and the totals were computed in code. Vendor A is a boutique that describes its method in writing, ties each tactic to a platform rule and offers two reference calls. Vendor B has the best reporting and the closest fit, but assures page-one positions within 60 days and ties part of its fee to that outcome. Vendor C is a larger firm with the strongest confidentiality practices and an unspectacular response elsewhere.

Criterion (weight)	Vendor A: score and points	Vendor B: score and points	Vendor C: score and points
Method and policy (25)	5 gives 25.0	1 gives 5.0	4 gives 20.0
Verifiable evidence (15)	3 gives 9.0	4 gives 12.0	4 gives 12.0
Fit to the matter (15)	4 gives 12.0	5 gives 15.0	3 gives 9.0
Confidentiality (15)	4 gives 12.0	3 gives 9.0	5 gives 15.0
Reporting (10)	3 gives 6.0	5 gives 10.0	3 gives 6.0
Commercial terms (10)	3 gives 6.0	2 gives 4.0	4 gives 8.0
Team (10)	4 gives 8.0	3 gives 6.0	3 gives 6.0
Total out of 100	78.0	61.0	76.0

Vendor B fails the gates: method scores 1, and the response assures an outcome and links its fee to it. Its total of 61.0 does not matter, and leading on fit and reporting does not rescue it. Had B scored 4 on method, its total would be 76.0, level with C. The gate, not the arithmetic, removed it.

Weighted totals, as a share of the top totalInvented example. Width is round(total / 78.0 * 100); labels show the total out of 100.

Vendor A78.0

Vendor C76.0

Vendor B, excluded by gate61.0

Totals computed from the example scores and weights above; no outside data. The widths are relative to the highest total, 78.0.

A and C are 2.0 points apart, which is not a result: one point of disagreement on a criterion weighted 15 moves a total by 3, and on one weighted 25 by 5. ## When scores are close: the tie rule and the sensitivity test A gap of less than 5 points out of 100 is a tie, and a tie is broken by evidence, not by a third scoring round. This is our working rule of thumb, based on how much one disagreement can move a total, and it is not a statistical threshold. Run a sensitivity test before you decide. Change the weights once, in a way a reasonable colleague might argue for, and see whether the order holds. In the example, moving method from 25 to 15, fit from 15 to 20 and reporting from 10 to 15, with the other weights unchanged, gives totals of 75.0 for A, 69.0 for B and 74.0 for C. The order of A, C and B is unchanged, and B moves much closer, which tells you its position depended on the heavy method weight and on the gate. If a reasonable change in weights flips first place, you do not have a winner yet. To break the tie between A and C, use three tools in order. Reference calls: ask each reference what the vendor refused to do, what it reported when a tactic failed, and how it handled a confidentiality concern. A paid first phase: buy a baseline audit of the same scope from both and score the output on the same sheet. An exit test: ask each vendor to describe in writing what happens to the work on day one after termination. The vendor that answers all three concretely wins. ## Next step: build the scorecard before you take any call Write the matter summary, the three gates and the seven weights on one page, and agree them with everyone who will score before any vendor speaks to you. Then get a baseline of where the subject stands, because it is the document every later report should be compared to.

If you are choosing a vendor now

Start with a baseline: what search engines and AI assistants show for the subject today, in writing. It becomes the reference for the RFP and for every report after. We offer a private audit for exactly this purpose: request an audit.

Reputation Partners sets out its approach on its ethics page, which gives you a comparison point for any vendor's own statement of method. ## Questions people ask ### How many vendors should receive the RFP? Three to five is usual. Fewer than three gives you no comparison, and more than five makes verification thin because reference calls and follow-up questions take time for each. Add a vendor only if it brings something the others lack, such as a specialism in your jurisdiction or in a type of content. Keep the same questions for every one. ### Should price be a scored criterion? Price sits inside "Commercial terms and exit" at 10 points, and it is better judged against scope than alone. A low fee for a vague scope scores low because you cannot tell what you are buying. Compare fees after scopes are made equal in the paid first phase, and treat a fee tied to an outcome as a gate matter, not a discount. ### Can I ask vendors for case studies? Yes, and score what comes back. Useful case material states the starting position, the actions, the dates and the outcome without naming a private client. Where a vendor cannot name anyone, ask for a reference who will take your call under the vendor's own confidentiality terms. Case studies with no dates or starting position score 1 on evidence. ### Do these rules apply outside the United States? The FTC rule is US law. The Google policies apply to Google Search and Google Maps wherever the user is. Other countries have their own rules on reviews, advertising and endorsements that were not read for this guide, so ask your counsel which apply to the subject and to the vendor, and put the answer in the RFP. ### What if the vendor I prefer fails a gate? Give it one written chance to restate the position, and score the restated answer. If it still assures an outcome or declines to describe its method, exclude it. A preferred vendor failing a gate is the situation the gate was written for, and the record of the decision protects you later.